Why does the WES catalogue advise that Jamicon WL-R miniature low ESR capacitors are "not suitable for audio"?
Here is the datasheet:
- Franc Zabkar
-- Please remove one 'i' from my address when replying by email.
:Why does the WES catalogue advise that Jamicon WL-R miniature low ESR :capacitors are "not suitable for audio"? : :Here is the datasheet: :
Franc, I see nothing on the WES data sheet or the manufacturer's data sheet
Looking at any of the Jamicon data sheets here
They're designed for circuits which put high frequency ripple current through them, so audiophiles will instantly notice an irritating 'shrill' element in the sound if they're used for audio.
Bob
PS - I was being sarcastic.
:On 6/02/2008 12:52 Ross Herbert wrote: :> On Wed, 06 Feb 2008 06:38:46 +1100, Franc Zabkar :> wrote: :> :> :Why does the WES catalogue advise that Jamicon WL-R miniature low ESR :> :capacitors are "not suitable for audio"? :> : :> :Here is the datasheet: :> :
Bob, I know that you wouldn't believe such things would result from using the WL in audio apps. However, it is just as well that you included your final remark otherwise someone from the golden ear brigade might just believe it was true.
WL
G'day Ross, Yeah, I thought I better make it clear that I wasn't serious, just in case someone believed it! :)
Cheers Bob
Maybe its because they aren't expensive enough ;) Obtaining and processing the snake oil required is very labour (left hand) intensive. :)
On Wed, 6 Feb 2008 15:16:03 -0800 (PST), kreed put finger to keyboard and composed:
I notice that elsewhere WES suggest that low leakage caps (ELR and EMRL) are the capacitors of choice for audio pre-amp stages. These have a leakage current of 0.3uA whereas the WL-R caps specify 3uA. However, the low impedance general purpose EXR caps with a leakage current of 3-4uA are stated to be OK for audio.
Of course over the page are "hifi" rated 10,000uF and 15,000uF caps, whatever that means, but we won't mention those. :-)
- Franc Zabkar
-- Please remove one 'i' from my address when replying by email.
On Feb 5, 8:38 pm, Franc Zabkar
Well, for example in ceramic capacitors with dielectrics like Z5U that is the same as Y5V, have the characteristic that the capacity is ONLY and ONLY achieved under bias. That meant a 1uF ceramic capacitor is only 100nF at 0.2V DC bias and ~1uF at 10V bias, etc. In audio applications you use capacitors in various functions. In power supply you don't care about such properties, because you always try to run at full rated voltage of the supply. However if you were to use such capacitor in oscillator or filters, you would get very large nonlinear distortion. I personally tried some small, cheapest ceramic capacitors (I think Z5U dielectric or similar) and I could tune oscillator frequancy depending on capacitor bias voltage with one finger on capacitor terminals and the other touching supply or ground. It was fun.
Now back to your question. Electrolytics should not have much nonlinearity at all. Maybe the double-layer capacitors before activation, or multi-electrolyte type formulations could have more exponential characteristic than is usual. But is is only a remote possibility. Maybe the question was whether you want also to use it as a bipolar non-biased capacitor. Soem 10 years ago I used electrolytic capacitor in an application which required only alternating current and of course the electrolytic capacitor - being only low current UNIPOLAR type smoked the electrolyte out. That is what are BIPOLAR electrolytic capacitor for. You ask WHAT is the difference? The main is the electrodes, being of aluminium foil, one is oxidized and one isn't. Bipolar capacitor has both electrodes oxidized (as far as I remember, weak, weak memories :-) ). Think about it briefly, would you use an electrolytic capacitor with reversed polarity? Of course not :). And that should be the most obvious matter on use of unipolar electrolytic capacitors in audio. Oh, forgot to mention what happens. Electrolytic capacitors have LIMITED lifetime. Wrong polarisation will result in much faster electrode corrosion and will greatly shorten the useful life. In audio path you want to use capacitors that will last and not change their parameters over time. In switched power supplies the capacitors have very short life, some will become useless in years and develop gas pressure from decomposition and heat inside. reversing polarity would much accelerate that.
If you want to use it under biased conditions of in a power supply, you don't need to worry about capacitor being suitable for SMPS or AUDIO, it only needs to be able deliver the current and not overheat during that.
On Wed, 20 Feb 2008 14:50:51 -0800 (PST), snipped-for-privacy@gmail.com put finger to keyboard and composed:
WL-R caps are polarised. The only time I can imagine a "bias" problem is when coupling the inputs and outputs of two items of audio equipment, but that would affect all non-polarised electrolytic caps, not just the subject ones.
- Franc Zabkar
-- Please remove one 'i' from my address when replying by email.
this is exactly backwards. 5% rated voltage gives about 10% drop in C.
50% rated voltage is more like 70% drop, full rated voltage closer to 90% (Z5U and Y5V are a bit different, but both are horrendous).in a given footprint I have yet to find a Y5V/Z5U cap that gives more capacitance than X7R when running at 70% rated voltage.
and their temperature behaviour is appalling, too.
I first "discovered" this circa 1993, when a SMPS output with a constant load had ~ 30x more ripple than I expected. after confirming the load, I went and looked hard at the cap, and noticed the voltage derating factor (25V cap, 20V supply). ISTR it was a 1uF Z5U, and I ended up with about
30nF at 20Vdc, 60C ambient. verified several ways: solder wires onto cap, stick on LCR bridge, heat with hot air - wow! add pullup resistor do DC supply, then put 20uF film cap in series with LCR bridge, and adjust V - wow! replace with 1uF film cap in circuit - ripple spot on. In the end I used a 220nF X7R and the ripple went *down* by 8xI have not used such a cap since then. And in fact specify in big letters on the sch "NO Z5U/Y5V caps" - in addition to specifying the dielectric in the BOM.
In audio
until of course you notice you havent got anything like as many farads as you thought.
However if you were to use such
X7R has about 15% drop in C at rated voltage IIRC, which makes it great for SMPS, but still s**te for these type of circuits. the tempco is similar, about 10% drop at Tmax.
they all have Barium Titanate & Titanium Dioxide in them (along with other goodies), which are very piezoelectric. whack your filter/osc with a pencil, and see the fun.
I have seriously thought about using them as $0.001 temperature sensors.....
it would be fun to make one into a microphone. although the sensor for an electronic drum would be easier - use 0.8mm PCB, and whack it with a drumstick....
NP0 is the best ceramic dielectric. ~ zero tempco, and flat with applied voltage. pretty much every smt cap below 1nF is NP0, but its hard to get them above 10nF or so. These are great for sensitive analog circuits.
I have seen a 100uF 200V NP0 capacitor with ultra-low ESL (~ 1nH IIRC). designed for smps use in satellites. and US$2/uF.
not at all - spot on.
Think about it briefly, would you
they look a bit like crappy diodes. hence getting unhappy and/or expiring.
And that should be the most obvious matter on use of unipolar
only in crappy smps. decent smps design is all about cap life (and thermal cycling of semiconductors).
alas almost all smps in consumer eqpt are crappy :(
indeed. nichicon & rubycon have some nice apps on figuring out how long a cap will live.
Cheers Terry
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.